Strategic interactions in competition policy: Dutch experiences1 Jarig van Sinderen2 Ron Kemp3 1 First draft. We are grateful to Jan Kees Winters and Erik Kloosterhuis for their helpful comments. 2 Chief Economist of the Netherlands Competition Authority, The Hague, The Netherlands. Correspondence address: [email protected] 3 Senior Member of Staff, Office of the Chief Economist, Netherlands Competition Authority, The Hague, The Netherlands. 1 Introduction It is well known that the Sherman Act of 1890 led to strategic behaviour by firms. For instance, since cartels were forbidden by the Sherman Act, the number of mergers increased in the USA. Especially after 1897 there was a clear and big increase in the number of mergers (Motta, 2004, pp. 5). This led to the Clayton Act of 1914 which (a.o.) established a legal foundation for merger control in the USA. These historical observations illustrate two points. First, the market reacts to institutional changes like changes in legislation, and, secondly, the legislator reacts to a new (market) situation with amendments to existing competition law. When competition legislation is introduced or amended in order to achieve a policy goal – which is generally supposed to be in the interest of the public- firms will either try to adhere to the law to their best possible advantage or to evade the law at least partly, depending on a cost-benefit analysis of both types of reaction. However, it takes some time before this process of adjusting to legislation fully settles down. This might explain why it took more than half a decade before the market reaction to the introduction of the Sherman Act became obvious. In a recent study by the Netherlands Bureau for Economic Policy Analyses (CPB) it is argued that in the period 1994-2001 the reaction to the introduction of the Netherlands competition law of 1998, took about two years. “The size of the coefficient of the lagged indicator suggests an adjustment of firms’ competitive behaviour to changes in the other determinants in the previous year. The coefficient of 0.44 implicates that competition is fully adjusted to initial changes of other determinants after two years.”(CPB, 2006: 37) So, changes in the law lead to reactions of economic agents but it takes time. This adjustment process may lead to a new equilibrium. In this equilibrium it is possible that the new law is fully adhered to, but it is also possible that it is partly absorbed with evasive behaviour. Full adherence would imply that cartels no longer exist, ‘abuse of dominance’ would no longer 2 occur and only efficient mergers would result. Full adherence would also imply that the enforcement of competition law would be flawless (no first- or second order mistakes); otherwise firms would adjust (arguably inefficiently) to these mistakes made by law enforcement. The legislative response to changes in market behaviour starts a process of action-reaction between the legislator and market parties. This process runs the risk that an abound of legislation and regulation is created as the legislator keeps adjusting the law to reach a ‘perfect’ outcome. The interaction between legislation and market parties calls for reconsideration of the law now and then. Regulation runs the risk of becoming obsolete, for instance. On the one hand it is necessary to update legislation. On the other hand competition policy which becomes too complicated because of regular adjustments might hamper the functioning of markets itself. The legislator should find a balance between effectiveness and efficiency in legislation. Apart from adapting to changes in law, there are also other kinds of reactions by market parties to existing competition law, competition authorities, the (competition) regulators and (private and public) law enforcement. We think of lobbying, lawsuits, public discussion, etc. In this paper, we will focus on brief descriptions of these interactions and (where possible) use anecdotal evidence of the Dutch context to illustrate them. As such, this paper may be viewed as the first step in trying to model analytically the relevant interactions in the context of competition policy. Alternatively, this way of proceeding might be viewed as a process of ‘proposition formation’4. The Netherlands’ competition law dates from 1998. It is essentially the same legislation as EU competition law. The law of 1998 substituted the law of 1956 called the WEM (Wet Economische Mededinging) which allowed cartel agreements as long as they were reported to the Ministry of Economic Affairs. Until recently in the Netherlands there were hardly any 4 It is a challenge to model these interactions, expectations formation and (out-of-equilibrium) behavioural adjustments in a full model. Even describing simple, yet realistic game-theoretical, models (in order to determine equilibrium outcomes) of the interactions between the three players (at the second level) is rather complicated. 3 changes in the Competition Act of 1998, but at this moment the first major changes in the law are in preparation, due to (for some part) certain market developments. From the start of the EU, competition law has been operative (though rigorously applied since the 1990’s only). Since EU competition law was stricter than the Dutch WEM was, a lot of Dutch competition cases in the period before 1998 were based on this EU law rather than on the WEM and handled by the European Commission rather than Dutch authorities.5 The remainder of this paper is structured as follows. First, we briefly elaborate on a framework to study the different types of interactions in the context of competition policy. Subsequently, we describe the different types of interactions in more detail in sections 3 to 6 using anecdotal evidence. We close with a discussion and implications for further research. 2 Framework for studying strategic interactions in competition policy In general there is an intricate pattern of actions and reactions between at least five types of player in the context of competition policy: the legislator, the competition authority (or regulator), the judiciary, firms, and the public at large (or consumers).6 We distinguish, for analytical purposes, two main (but related) levels of interactions. - The first level concerns the ultimate effect of legislation on market outcomes. To put it simply: the legislator introduces competition law and a competition authority,7 given certain market outcomes (for instance, as was the case in the Netherlands, market outcomes that were determined in a possibly large number of markets by cartels) and this leads (over time) to (probably) different market outcomes. These market outcomes might either be part of an 5 A famous example is the 1992-Commission Decision on Bid Rigging in the Construction sector in the Netherlands (European Commission, 1992, Building and Construction in the Netherlands, 05.02.1992, Official Journal L92 – 7/04/1992, page 1, celex 392D0204. See also Russo, Carree, Günster and Schinkel (forthcoming). 6 For the remainder we will focus on the first four types of players. Only as far as ‘private litigation’ is concerned, we will discuss the role of the consumer. 7 At this conceptual level one may also read this statement as: the legislator introduces a specifically regulatory framework, including a regulator. We will therefore speak of ‘regulator’ or ‘competition authority’ as ‘enforcement agency’. 4 adjustment process over time, or they might be new equilibrium outcomes. The outcomes are determined by the interactions between the other players (enforcement agencies, judiciary, firms). In that sense, those interactions are a ‘black box’ at this level of analysis. Whichever will be the case; these outcomes of the ‘black box’ may be considered unsatisfactory by the legislator and have an incentive to change the law in order to remedy the problems as conceived by the legislator. The discontent of the legislator may be based on information or complaints of the other players, for instances by means of lobbying activities. Examples of this will be described below in more detail, but one can think of changes in the Netherlands’ competition law concerning the discussion on the personal liability of managers on top of liability of firms. Changes in the law are also partly due to exogenous reasons, like the impact of European rules.8 - The second level concerns the closer scrutiny of the first level’s ‘black box’. What are the interactions between enforcement agency, the judiciary and the firms and why might the outcomes be considered undesirable (and in what sense), so that changes in the law might be required? These interactions determine the actual market-outcomes and whether or not these are considered to be (un)satisfactory. Therefore it is very useful to look into these processes in more detail. Adjustment of the rules of interaction or strategic behaviour to influence these interactions and the outcome can prevent changes in the law and will improve the effectiveness of legislation. These interactions, as stated before, are necessarily of a dynamic nature. The introduction of competition law and its’ enforcement induce reactions by firms and the judiciary. These reactions lead to yet other reactions by the competition authority, and so on. Expectations are formed and adjusted over time until some (new) equilibrium might be reached. Still, the possibility exists that during the adjustment process, out-of-equilibrium market outcomes are so that they are considered to be unsatisfactory and will lead to adjustments in either the law or in the application of enforcement. To study the interactions at the second level, we will distinguish five types of interactions (for ease of exposition). The first type of interaction has to do with changes that are induced by the introduction of the competition act. The second type of interaction reflects some sort of an 8 Those interactions will not be described here, though. 5 equilibrium situation in which expectations about the enforcement agency’s activity have been formed (and the enforcement agency has to decide how to make use of this). The third type of interaction is about type I and type II-errors (i.e. ‘false prosecutions’ and ‘false non- prosecutions’) and their influence on behaviour. The fourth type of interaction considers the role of the judiciary and the last one is about private litigation. A) At the introduction of the competition act: - firms establish expectations about enforcement and actions taken; - the enforcement agency establish expectations about firms’ behaviour and actions taken; - based on the actions and expectations of both parties, there can be adjustments in behaviour. This process will be described in section 3. Topics we discuss are: the enormous amount of exemption requests by market parties after the introduction of the competition act, the thoroughness which the NMa displayed in treating the exemption requests in order to gain a reputation, but which also had some negative consequences, for example on the pro-active detection of cartels or other abuses. B) At any later stage in which expectations have been set (more or less): - a possible change of behaviour by the enforcement agency may occur, and/or; - a possible change of behaviour by firms may occur. This will be described in section 4. We will discuss the use of alternative ways of enforcement and the influence of ‘leniency’ and ‘fining’ on cartel formation specifically. We also present the results from a recent research paper on the effects of the Netherlands Competition authority (hereafter NMa) as anticipated by companies. This longitudinal research shows that companies increasingly take the competition law and the competition law enforcement into account in their business decisions. C) At any later stage in which (possibly) type I and type II errors become more or less clear:9 - the enforcement agency may change their behaviour in enforcement activities and decisions; 9 Typically the judiciary will play an active role in the establishment of both types, but we will not focus on the interactions by all three players. The role of the judiciary will be separately dealt with. 6 - how firms react to these types of errors. This will be described in section 5. We will discuss the possible problems that firms might experience as a consequence of these type of errors. For example, companies might wish to sue other companies in order to ‘raise rivals’ costs’. Opmerking [r1]: Is dit relevant in deze context (heeft dit met type 1 en 2 te maken) of moeten we hier bv Nuon Reliant opnemen, fusie doorlaten gaan met D) At the stage where the judiciary shows its’ influence: remedies maar daarna in beroep om dit terug te draaien en schade te claimen. - the judiciary will have an impact on the enforcement agency’s actions, both with regard to the judiciary and the firms; - the judiciary will influence the firms’ expectations and actions taken; - the decisions of the judiciary will also have an impact on private litigation and, hence, on firms’ behaviour. The judiciary interferes with the effectiveness of the enforcement agency. This influence can be positive or negative, but we will only deal with the judiciary as a determinant for type I and type II-errors in the sense that lawsuits are being used as a strategic element by firms (hence also in section 5). E) Private litigation: - firms may have the incentives to litigate and this will have consequences on the market outcomes; - (groups of) consumers may have an incentives to litigate which also has influence on the market outcome10. This will be described in section 6. Although an important topic, because it might be one way of filling the possible gap in the enforcement agency’s effectiveness,11 it does not as yet have an important role to play in the Netherlands. 3 Reactions to the introduction of the Competition Act The Netherlands used to be called a cartel-paradise. Before 1998 cartels were allowed when they were registered, hence the new law, that prohibited formerly allowed behaviour, was 10 In this case we will not discuss the possible reactions of the enforcement agency to these private litigations. An example of such a reaction might be that the competition authority will fine less severely, which will lead to different reations by the other players involved. 11 Alternatively, if effective, private litigation might also sustain the enforcement agency’s policy. 7 quite a change to firms and enforcement. The NMa needed to establish a reputation as a real ‘cartel-fighter’, firms would try to resist the implications of the law. We will discuss some elements of the adjustment process, i.e. the uneasiness with the end of the cartel paradise, the exemptions asked by firms as a consequence and the (re)actions by the NMa. Uneasiness with the end of the cartel paradise The competition law of the 1956 (WEM) allowed companies to make agreements on price policies, on geographical market division, etc. The only condition to allow such a practice was that companies had to report their agreement to the Ministry of Economic Affairs, which registered the agreement in a register (the Dutch “Kartel Register”). Sectors which were used to cartelise may have a longer process to adapt to the new situation. In our opinion it is not a surprise that the first illegal agreements after the fundamental change in the competition law were found in sectors which in the old days were allowed to form a cartel One can think of the construction sector in the Netherlands and other sectors that had difficulties to relinquish previous price agreements after the introduction of the competition law. The long standing tradition of cartel behaviour might have led to a very slow adjustment process. Also the enormous amount of exemption requests by firms at the introduction of the competition act, which have taken years to be assessed by the NMa is an example of such an adjustment process . Exemption The number of exemption application was about 1.100 in the first two years after the competition law came into force. This number of request was considerably higher than expected. Of these requests, 47% of them turned out not to violate any of the new competition rules. Of the other 53%, only 9% were granted as exemption. These unnecessary applications can be considered as adjustment costs to the new competition law.12 A limited number of the actual cases which were potentially in conflict with the new competition law were granted an exemption. In our opinion this case illustrates that companies were not aware of the content of the Act. They asked for exemptions for activities which were not in conflict with the new 12 According to Ministry of Economic Affairs (2003) the administrative burdens of the exemption applications were about € 600.000, about 29% of the total administrative burden of the Competition law. 8 competition law. Other companies were used to the habit of price agreements, bid rigging and/or a division of markets and had the impression that these practices could be continued under the new law. This case shows that it is very important to give good information on laws and changes in it in order to avoid adjustment costs as much as possible. Role of the NMa The NMa had an important role to play, but was of course inexperienced when it was newly established. The NMa needed to prove itself to be an effective enforcer, but was confronted with the large number of requests for an exemption. Therefore, the NMa started very thorough assessments of these requests. This lead to positive spin-offs, like interesting information about Dutch economic sectors and knowledge about the application of competition law in practice. It also gave the NMa the opportunity to inform firms and sectors about the limits of competition law. There was a price to be paid, though. There was relatively little time (in 1998-2000) for other enforcement actions, such as pro-active cartel investigations. This, of course, may have contributed to a longer period of time before the NMa was considered by firms to be a vigorous cartel-fighter. In the next section, we will discuss in more detail how these perceptions work and how the firms’ perceptions about the NMa develop over time. Changes in the law One of the main changes in the Competition Act that are currently being discussed in Parliament is to make managers personally liable for violations of the Competition Act. One of the consequences is that they can be fined, but the possibility of imprisonment is also discussed. An important reason for this discussion is to increase the deterrent effect of the Competition Act. Another, related, change in the Competition Act is the possibility to search not only a firm’s premises, but also those of the managers. Also this is a reaction to anticipation by managers and forms on NMa cartel investigations. There were some cases where there is evidence that papers on cartel agreements were kept on private premises. Lobbying and shelter for special interest groups to influence the scope of the competition law Lobbying activities often focus on regulation for a particular group instead of the competition law in general. For instance, professionals who operate under the protection of a specific law (exemptions on the Competition law) tend to have a strong position they want to protect. Political discussions on the introduction of a more market oriented change is avoided and often blocked by pressure groups. Lobbying is a very strong instrument in order to influence 9 the policymaker in favour of protection of a certain profession. “Famous” examples in the Netherlands are the “public notaries”. Although more competition has been introduced some politicians are afraid that the quality of the notary services will suffer as a consequence of a more intense competition.13 The notary does no have enough incentives to oppose such a political uneasiness. The opposition against more competition in these markets is often based on the argument that quality control is of the utmost interest of the public and can not be guaranteed without government intervention. These arguments are not supported by a lot of professionals. There are, however, also lobby activities related to the Competition law. Representation organisations of small retailers lobbied for an adjustment of the Competition law. They wanted to include a prohibition in the Competition law to sell below the purchasing costs as a response to the intense competition between larger supermarket chains. They claimed that the intense competition dispel smaller retailers from the market. Recently, the Dutch Electricity sector lobbied intensively to prevent the split of their companies into a commercial distribution part and a regulated network part. Although the Minister of Economic Affairs and the House of Representatives were in favour of a new law to split the companies, the companies persuaded the Senate to abandon the split, at least initially. 4 Expectations and behaviour Stuk over leniency zoals aangegeven onder punt B?? An important way to enforce the competition law is voluntary compliance by companies. This anticipation effect is very important as a large majority of companies will comply as a small number of the companies will be involved in unlawful behaviour. The question is, however, how to influence this anticipation effect? To understand this, it is important to understand managerial decision making. Managerial decisions, such as those related to competition, are based upon simplifying and framing processes. Managers use mental models as simplifications of reality (Chernatorty, Daniels and Johnson, 1993). Based on observations and expectations about competitive forces, managers develop their perception about the 13 This is a paradox. Economist think that competition might lead to higher quality of services for a lower price, while politicians claim that competition might lead to a lower price and lower quality. 10
Description: